Alteration by natural processes or anthropogenic manipulation? Assessing human skull breakage through machine learning algorithms

• DOI: 10.1007/s12520-024-02083-5
• Publication Date: 2024-10-11T11:01:46Z
• Authors: Francesc Marginedas; Abel Moclán; Miriam Cubas; Asier Gómez-Olive...; Palmira Saladié; Antonio Rodríguez...
• Keywords: Breakage

Abstract Bone breakage is one of the most common features in archaeological record. Fractures occur at different times and are classified as fresh or dry depending on presence absence collagen bone. In study human remains, timing occurrence a fracture crucial importance it can sometimes be linked to cause death. Types skull based when they occurred, though not all fractures correspond expected features. This variability added challenge working with bones covered consolidant, which obstructs bone surface hinders taphonomic analysis. case Txispiri calotte, was categorized cup early 20th century, this classification later rejected 1990s. study, we used statistics machine learning (ML) test characteristics set fragments fractures, another calotte. For purpose, considered type, trajectory, angles, cortical delamination texture each individual fractures. Our results show that 13 calotte bear no relation artificially produced cups. shows potential ML algorithms classify within same specimen, method applied other assemblages similar characteristics.